diff --git a/docs/deploy/agent-runtime/deploy.md b/docs/deploy/agent-runtime/deploy.md index aacefb6512..4e7b888afe 100644 --- a/docs/deploy/agent-runtime/deploy.md +++ b/docs/deploy/agent-runtime/deploy.md @@ -9,13 +9,13 @@ ADK agent code to Google Cloud [Agent Runtime](https://cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/overview). You should follow this deployment path if you have an existing Google Cloud project and if you want to carefully manage deploying an ADK agent to Agent -Engine runtime environment. These instructions use Cloud Console, the gcloud +Runtime environment. These instructions use Cloud Console, the gcloud command line interface, and the ADK command line interface (ADK CLI). This path is recommended for users who are already familiar with configuring Google Cloud projects, and users preparing for production deployments. These instructions describe how to deploy an ADK project to Google Cloud Agent -Engine runtime environment, which includes the following stages: +Runtime environment, which includes the following stages: * [Setup Google Cloud project](#setup-cloud-project) * [Prepare agent project folder](#define-your-agent) diff --git a/docs/deploy/agent-runtime/test.md b/docs/deploy/agent-runtime/test.md index df5fb13cf4..ff6b947536 100644 --- a/docs/deploy/agent-runtime/test.md +++ b/docs/deploy/agent-runtime/test.md @@ -289,7 +289,7 @@ async for event in remote_app.async_stream_query( If you have performed deployments as tests, it is a good practice to clean up your cloud resources after you have finished. You can delete the deployed Agent -Engine instance to avoid any unexpected charges on your Google Cloud account. +Runtime instance to avoid any unexpected charges on your Google Cloud account. ```python remote_app.delete(force=True) diff --git a/docs/grounding/index.md b/docs/grounding/index.md index 4509c443a2..549ec21e27 100644 --- a/docs/grounding/index.md +++ b/docs/grounding/index.md @@ -6,7 +6,7 @@ ADK supports multiple grounding approaches: - **Google Search Grounding**: Connect agents to real-time web information for queries requiring current data like news, weather, or facts that may have changed since the model's training. - **Grounding with Search**: Connect agents to your organization's private documents and enterprise data for queries requiring proprietary information. -- **Agentic RAG**: Build agents that reason about how to search, constructing queries and filters dynamically using Vector Search 2.0, RAG Engine, or other retrieval systems. +- **Agentic RAG**: Build agents that reason about how to search, constructing queries and filters dynamically using Agent Retrieval, Knowledge Engine, or other retrieval systems.
@@ -54,7 +54,7 @@ ADK supports multiple grounding approaches: --- - A document Q&A agent powered by RAG Engine. Upload documents and ask questions to receive accurate answers with citations formatted as URLs pointing to source materials. + A document Q&A agent powered by Knowledge Engine. Upload documents and ask questions to receive accurate answers with citations formatted as URLs pointing to source materials. - [RAG Agent](https://github.com/google/adk-samples/tree/main/python/agents/RAG) diff --git a/docs/integrations/adspirer.md b/docs/integrations/adspirer.md index 8aed2bab23..f8ac61973d 100644 --- a/docs/integrations/adspirer.md +++ b/docs/integrations/adspirer.md @@ -88,7 +88,7 @@ agent can operate autonomously with built-in protections. from mcp import StdioServerParameters root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="advertising_agent", instruction=( "You are an advertising agent that helps users create, manage, " @@ -125,7 +125,7 @@ agent can operate autonomously with built-in protections. ADSPIRER_ACCESS_TOKEN = "YOUR_ADSPIRER_ACCESS_TOKEN" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="advertising_agent", instruction=( "You are an advertising agent that helps users create, manage, " @@ -157,7 +157,7 @@ agent can operate autonomously with built-in protections. import { LlmAgent, MCPToolset } from "@google/adk"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "advertising_agent", instruction: "You are an advertising agent that helps users create, manage, " + @@ -192,7 +192,7 @@ agent can operate autonomously with built-in protections. const ADSPIRER_ACCESS_TOKEN = "YOUR_ADSPIRER_ACCESS_TOKEN"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "advertising_agent", instruction: "You are an advertising agent that helps users create, manage, " + diff --git a/docs/integrations/agentmail.md b/docs/integrations/agentmail.md index 8071da0ede..2ba6d0c2ef 100644 --- a/docs/integrations/agentmail.md +++ b/docs/integrations/agentmail.md @@ -52,7 +52,7 @@ language. AGENTMAIL_API_KEY = "YOUR_AGENTMAIL_API_KEY" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="agentmail_agent", instruction="Help users manage email inboxes and send messages", tools=[ @@ -85,7 +85,7 @@ language. const AGENTMAIL_API_KEY = "YOUR_AGENTMAIL_API_KEY"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "agentmail_agent", instruction: "Help users manage email inboxes and send messages", tools: [ diff --git a/docs/integrations/agentphone.md b/docs/integrations/agentphone.md index 30d96877b8..4f26f15b94 100644 --- a/docs/integrations/agentphone.md +++ b/docs/integrations/agentphone.md @@ -59,7 +59,7 @@ create autonomous AI voice agents using natural language. AGENTPHONE_API_KEY = "YOUR_AGENTPHONE_API_KEY" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="agentphone_agent", instruction="Help users make phone calls, send SMS, and manage phone numbers", tools=[ @@ -92,7 +92,7 @@ create autonomous AI voice agents using natural language. AGENTPHONE_API_KEY = "YOUR_AGENTPHONE_API_KEY" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="agentphone_agent", instruction="Help users make phone calls, send SMS, and manage phone numbers", tools=[ @@ -118,7 +118,7 @@ create autonomous AI voice agents using natural language. const AGENTPHONE_API_KEY = "YOUR_AGENTPHONE_API_KEY"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "agentphone_agent", instruction: "Help users make phone calls, send SMS, and manage phone numbers", tools: [ @@ -146,7 +146,7 @@ create autonomous AI voice agents using natural language. const AGENTPHONE_API_KEY = "YOUR_AGENTPHONE_API_KEY"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "agentphone_agent", instruction: "Help users make phone calls, send SMS, and manage phone numbers", tools: [ diff --git a/docs/integrations/asana.md b/docs/integrations/asana.md index 797842d0c6..1cde233ffe 100644 --- a/docs/integrations/asana.md +++ b/docs/integrations/asana.md @@ -44,7 +44,7 @@ tasks, goals, and team collaboration using natural language. from mcp import StdioServerParameters root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="asana_agent", instruction="Help users manage projects, tasks, and goals in Asana", tools=[ @@ -73,7 +73,7 @@ tasks, goals, and team collaboration using natural language. import { LlmAgent, MCPToolset } from "@google/adk"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "asana_agent", instruction: "Help users manage projects, tasks, and goals in Asana", tools: [ diff --git a/docs/integrations/atlassian.md b/docs/integrations/atlassian.md index 147ae6d5ee..6868e90747 100644 --- a/docs/integrations/atlassian.md +++ b/docs/integrations/atlassian.md @@ -48,7 +48,7 @@ collaboration workflows using natural language. root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="atlassian_agent", instruction="Help users work with data in Atlassian products", tools=[ @@ -77,7 +77,7 @@ collaboration workflows using natural language. import { LlmAgent, MCPToolset } from "@google/adk"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "atlassian_agent", instruction: "Help users work with data in Atlassian products", tools: [ diff --git a/docs/integrations/cartesia.md b/docs/integrations/cartesia.md index 9dda495b4c..594ebb533d 100644 --- a/docs/integrations/cartesia.md +++ b/docs/integrations/cartesia.md @@ -53,7 +53,7 @@ across languages, and create audio content using natural language. CARTESIA_API_KEY = "YOUR_CARTESIA_API_KEY" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="cartesia_agent", instruction="Help users generate speech and work with audio content", tools=[ @@ -84,7 +84,7 @@ across languages, and create audio content using natural language. const CARTESIA_API_KEY = "YOUR_CARTESIA_API_KEY"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "cartesia_agent", instruction: "Help users generate speech and work with audio content", tools: [ diff --git a/docs/integrations/chroma.md b/docs/integrations/chroma.md index d1da42935f..5dace327d1 100644 --- a/docs/integrations/chroma.md +++ b/docs/integrations/chroma.md @@ -56,7 +56,7 @@ search, and metadata filtering. # CHROMA_API_KEY = "your-api-key" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="chroma_agent", instruction="Help users store and retrieve information using semantic search", tools=[ @@ -105,7 +105,7 @@ search, and metadata filtering. // const CHROMA_API_KEY = "your-api-key"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "chroma_agent", instruction: "Help users store and retrieve information using semantic search", tools: [ diff --git a/docs/integrations/code-exec-agent-runtime.md b/docs/integrations/code-exec-agent-runtime.md index 52980bccc4..943b7c78f3 100644 --- a/docs/integrations/code-exec-agent-runtime.md +++ b/docs/integrations/code-exec-agent-runtime.md @@ -13,7 +13,7 @@ catalog_tags: ["code", "google"] The Agent Runtime Code Execution ADK Tool provides a low-latency, highly efficient method for running AI-generated code using the -[Google Cloud Agent Engine](https://cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/overview) +[Google Cloud Agent Runtime](https://cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/overview) service. This tool is designed for fast execution, tailored for agentic workflows, and uses sandboxed environments for improved security. The Code Execution tool allows code and data to persist over multiple requests, enabling complex, @@ -28,7 +28,7 @@ This code execution tool is part of the Agent Runtime suite, however you do not have to deploy your agent to Agent Runtime to use it. You can run your agent locally or with other services and use this tool. For more information about the Code Execution feature in Agent Runtime, see the -[Agent Engine Code Execution](https://cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/code-execution/overview) +[Agent Runtime Code Execution](https://cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/code-execution/overview) documentation. diff --git a/docs/integrations/couchbase.md b/docs/integrations/couchbase.md index 7ed7db0246..1521f161d8 100644 --- a/docs/integrations/couchbase.md +++ b/docs/integrations/couchbase.md @@ -56,7 +56,7 @@ issues. CB_PASSWORD = "password" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="couchbase_agent", instruction="Help users explore and query Couchbase databases", tools=[ @@ -91,7 +91,7 @@ issues. const CB_PASSWORD = "password"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "couchbase_agent", instruction: "Help users explore and query Couchbase databases", tools: [ diff --git a/docs/integrations/daytona.md b/docs/integrations/daytona.md index 25ef16e10a..3c93230166 100644 --- a/docs/integrations/daytona.md +++ b/docs/integrations/daytona.md @@ -50,7 +50,7 @@ plugin = DaytonaPlugin( ) root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="sandbox_agent", instruction="Help users execute code and commands in a secure sandbox", tools=plugin.get_tools(), diff --git a/docs/integrations/elevenlabs.md b/docs/integrations/elevenlabs.md index 2cb33dd998..a528733bd6 100644 --- a/docs/integrations/elevenlabs.md +++ b/docs/integrations/elevenlabs.md @@ -57,7 +57,7 @@ AI experiences using natural language. ELEVENLABS_API_KEY = "YOUR_ELEVENLABS_API_KEY" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="elevenlabs_agent", instruction="Help users generate speech, clone voices, and process audio", tools=[ @@ -87,7 +87,7 @@ AI experiences using natural language. const ELEVENLABS_API_KEY = "YOUR_ELEVENLABS_API_KEY"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "elevenlabs_agent", instruction: "Help users generate speech, clone voices, and process audio", tools: [ diff --git a/docs/integrations/firestore-session-service.md b/docs/integrations/firestore-session-service.md index a3df380c31..469df777a3 100644 --- a/docs/integrations/firestore-session-service.md +++ b/docs/integrations/firestore-session-service.md @@ -154,7 +154,7 @@ public class YourAgentApplication { You are a helpful assistant that tells the current time in a city. Use the 'getCurrentTime' tool for this purpose. \""") - .model("gemini-3.1-pro-preview") + .model("gemini-flash-latest") .tools(FunctionTool.create(YourAgentApplication.class, "getCurrentTime")) .build(); } diff --git a/docs/integrations/galileo.md b/docs/integrations/galileo.md index 4db9b2647b..487829edb5 100644 --- a/docs/integrations/galileo.md +++ b/docs/integrations/galileo.md @@ -94,7 +94,7 @@ def get_current_time(city: str) -> dict: root_agent = Agent( - model="gemini-3-flash-preview", + model="gemini-flash-latest", name="root_agent", description="Tells the current time in a specified city.", instruction=( diff --git a/docs/integrations/github.md b/docs/integrations/github.md index 9541e5460a..407095e64e 100644 --- a/docs/integrations/github.md +++ b/docs/integrations/github.md @@ -46,7 +46,7 @@ automate workflows using natural language. GITHUB_TOKEN = "YOUR_GITHUB_TOKEN" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="github_agent", instruction="Help users get information from GitHub", tools=[ @@ -74,7 +74,7 @@ automate workflows using natural language. const GITHUB_TOKEN = "YOUR_GITHUB_TOKEN"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "github_agent", instruction: "Help users get information from GitHub", tools: [ diff --git a/docs/integrations/gitlab.md b/docs/integrations/gitlab.md index 378e00f97f..80873be0bc 100644 --- a/docs/integrations/gitlab.md +++ b/docs/integrations/gitlab.md @@ -56,7 +56,7 @@ searches, and automate development workflows using natural language. GITLAB_INSTANCE_URL = "gitlab.com" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="gitlab_agent", instruction="Help users get information from GitLab", tools=[ @@ -90,7 +90,7 @@ searches, and automate development workflows using natural language. const GITLAB_INSTANCE_URL = "gitlab.com"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "gitlab_agent", instruction: "Help users get information from GitLab", tools: [ diff --git a/docs/integrations/google-developer-knowledge.md b/docs/integrations/google-developer-knowledge.md index 85fdc3de2e..e44ed223e1 100644 --- a/docs/integrations/google-developer-knowledge.md +++ b/docs/integrations/google-developer-knowledge.md @@ -52,7 +52,7 @@ Guide](https://developers.google.com/knowledge/mcp#installation) for the precise DEVELOPER_KNOWLEDGE_API_KEY = "YOUR_DEVELOPER_KNOWLEDGE_API_KEY" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="google_knowledge_agent", instruction="Search Google developer documentation for implementation guidance.", tools=[ @@ -76,7 +76,7 @@ Guide](https://developers.google.com/knowledge/mcp#installation) for the precise const DEVELOPER_KNOWLEDGE_API_KEY = "YOUR_DEVELOPER_KNOWLEDGE_API_KEY"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "google_knowledge_agent", instruction: "Search Google developer documentation for implementation guidance.", tools: [ diff --git a/docs/integrations/hugging-face.md b/docs/integrations/hugging-face.md index 4d1bccc5c2..a2cc687779 100644 --- a/docs/integrations/hugging-face.md +++ b/docs/integrations/hugging-face.md @@ -45,7 +45,7 @@ your ADK agent to the Hugging Face Hub and thousands of Gradio AI Applications. HUGGING_FACE_TOKEN = "YOUR_HUGGING_FACE_TOKEN" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="hugging_face_agent", instruction="Help users get information from Hugging Face", tools=[ @@ -78,7 +78,7 @@ your ADK agent to the Hugging Face Hub and thousands of Gradio AI Applications. HUGGING_FACE_TOKEN = "YOUR_HUGGING_FACE_TOKEN" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="hugging_face_agent", instruction="Help users get information from Hugging Face", tools=[ @@ -104,7 +104,7 @@ your ADK agent to the Hugging Face Hub and thousands of Gradio AI Applications. const HUGGING_FACE_TOKEN = "YOUR_HUGGING_FACE_TOKEN"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "hugging_face_agent", instruction: "Help users get information from Hugging Face", tools: [ @@ -132,7 +132,7 @@ your ADK agent to the Hugging Face Hub and thousands of Gradio AI Applications. const HUGGING_FACE_TOKEN = "YOUR_HUGGING_FACE_TOKEN"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "hugging_face_agent", instruction: "Help users get information from Hugging Face", tools: [ diff --git a/docs/integrations/rag-engine.md b/docs/integrations/knowledge-engine.md similarity index 55% rename from docs/integrations/rag-engine.md rename to docs/integrations/knowledge-engine.md index 0a8b7cf4fe..ee649ee266 100644 --- a/docs/integrations/rag-engine.md +++ b/docs/integrations/knowledge-engine.md @@ -1,20 +1,25 @@ --- -catalog_title: RAG Engine -catalog_description: Perform private data retrieval using RAG Engine +catalog_title: Knowledge Engine +catalog_description: Perform private data retrieval using Knowledge Engine catalog_icon: /integrations/assets/agent-platform.svg catalog_tags: ["data","google"] --- -# RAG Engine tool for ADK +# Knowledge Engine tool for ADK
Supported in ADKPython v0.1.0Java v0.2.0
-The `vertex_ai_rag_retrieval` tool allows the agent to perform private data retrieval using RAG Engine. +The `vertex_ai_rag_retrieval` tool allows the agent to perform private data +retrieval using Knowledge Engine. -When you use grounding with RAG Engine, you need to prepare a RAG corpus beforehand. -Please refer to the [RAG ADK agent sample](https://github.com/google/adk-samples/blob/main/python/agents/RAG/rag/shared_libraries/prepare_corpus_and_data.py) or [RAG Engine page](https://cloud.google.com/vertex-ai/generative-ai/docs/rag-engine/rag-quickstart) for setting it up. +When you use grounding with Knowledge Engine, you need to prepare a RAG corpus +beforehand. Please refer to the [RAG ADK agent +sample](https://github.com/google/adk-samples/blob/main/python/agents/RAG/rag/shared_libraries/prepare_corpus_and_data.py) +or [Knowledge Engine +page](https://cloud.google.com/vertex-ai/generative-ai/docs/rag-engine/rag-quickstart) +for setting it up. !!! warning "Warning: Single tool per agent limitation" diff --git a/docs/integrations/linear.md b/docs/integrations/linear.md index c6e608b41c..458f8ab031 100644 --- a/docs/integrations/linear.md +++ b/docs/integrations/linear.md @@ -50,7 +50,7 @@ project cycles, and automate development workflows using natural language. from mcp import StdioServerParameters root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="linear_agent", instruction="Help users manage issues, projects, and cycles in Linear", tools=[ @@ -88,7 +88,7 @@ project cycles, and automate development workflows using natural language. LINEAR_API_KEY = "YOUR_LINEAR_API_KEY" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="linear_agent", instruction="Help users manage issues, projects, and cycles in Linear", tools=[ @@ -118,7 +118,7 @@ project cycles, and automate development workflows using natural language. import { LlmAgent, MCPToolset } from "@google/adk"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "linear_agent", instruction: "Help users manage issues, projects, and cycles in Linear", tools: [ @@ -150,7 +150,7 @@ project cycles, and automate development workflows using natural language. const LINEAR_API_KEY = "YOUR_LINEAR_API_KEY"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "linear_agent", instruction: "Help users manage issues, projects, and cycles in Linear", tools: [ diff --git a/docs/integrations/mailgun.md b/docs/integrations/mailgun.md index 3cf4ae7ff0..fed76d4c57 100644 --- a/docs/integrations/mailgun.md +++ b/docs/integrations/mailgun.md @@ -49,7 +49,7 @@ natural language. MAILGUN_API_KEY = "YOUR_MAILGUN_API_KEY" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="mailgun_agent", instruction="Help users send emails and manage their Mailgun account", tools=[ @@ -83,7 +83,7 @@ natural language. const MAILGUN_API_KEY = "YOUR_MAILGUN_API_KEY"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "mailgun_agent", instruction: "Help users send emails and manage their Mailgun account", tools: [ diff --git a/docs/integrations/mongodb.md b/docs/integrations/mongodb.md index 421b3e85df..b8f7e1e4b6 100644 --- a/docs/integrations/mongodb.md +++ b/docs/integrations/mongodb.md @@ -57,7 +57,7 @@ using natural language. # ATLAS_CLIENT_SECRET = "YOUR_ATLAS_CLIENT_SECRET" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="mongodb_agent", instruction="Help users query and manage MongoDB databases", tools=[ @@ -100,7 +100,7 @@ using natural language. // const ATLAS_CLIENT_SECRET = "YOUR_ATLAS_CLIENT_SECRET"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "mongodb_agent", instruction: "Help users query and manage MongoDB databases", tools: [ diff --git a/docs/integrations/n8n.md b/docs/integrations/n8n.md index 5f5769e3aa..4a099507d5 100644 --- a/docs/integrations/n8n.md +++ b/docs/integrations/n8n.md @@ -69,7 +69,7 @@ for detailed setup instructions. N8N_MCP_TOKEN = "YOUR_N8N_MCP_TOKEN" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="n8n_agent", instruction="Help users manage and execute workflows in n8n", tools=[ @@ -104,7 +104,7 @@ for detailed setup instructions. N8N_MCP_TOKEN = "YOUR_N8N_MCP_TOKEN" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="n8n_agent", instruction="Help users manage and execute workflows in n8n", tools=[ @@ -131,7 +131,7 @@ for detailed setup instructions. const N8N_MCP_TOKEN = "YOUR_N8N_MCP_TOKEN"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "n8n_agent", instruction: "Help users manage and execute workflows in n8n", tools: [ @@ -164,7 +164,7 @@ for detailed setup instructions. const N8N_MCP_TOKEN = "YOUR_N8N_MCP_TOKEN"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "n8n_agent", instruction: "Help users manage and execute workflows in n8n", tools: [ diff --git a/docs/integrations/notion.md b/docs/integrations/notion.md index 98ff68096a..7f67ee1432 100644 --- a/docs/integrations/notion.md +++ b/docs/integrations/notion.md @@ -58,7 +58,7 @@ language. NOTION_TOKEN = "YOUR_NOTION_TOKEN" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="notion_agent", instruction="Help users get information from Notion", tools=[ @@ -91,7 +91,7 @@ language. const NOTION_TOKEN = "YOUR_NOTION_TOKEN"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "notion_agent", instruction: "Help users get information from Notion", tools: [ diff --git a/docs/integrations/paypal.md b/docs/integrations/paypal.md index f87651c9ef..49db3cad73 100644 --- a/docs/integrations/paypal.md +++ b/docs/integrations/paypal.md @@ -55,7 +55,7 @@ workflows and business insights. PAYPAL_ACCESS_TOKEN = "YOUR_PAYPAL_ACCESS_TOKEN" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="paypal_agent", instruction="Help users manage their PayPal account", tools=[ @@ -93,7 +93,7 @@ workflows and business insights. PAYPAL_ACCESS_TOKEN = "YOUR_PAYPAL_ACCESS_TOKEN" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="paypal_agent", instruction="Help users manage their PayPal account", tools=[ @@ -120,7 +120,7 @@ workflows and business insights. const PAYPAL_ACCESS_TOKEN = "YOUR_PAYPAL_ACCESS_TOKEN"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "paypal_agent", instruction: "Help users manage their PayPal account", tools: [ diff --git a/docs/integrations/pinecone.md b/docs/integrations/pinecone.md index f4171a09a7..ed7b3912e5 100644 --- a/docs/integrations/pinecone.md +++ b/docs/integrations/pinecone.md @@ -48,7 +48,7 @@ filtering, and search across multiple indexes with reranking. PINECONE_API_KEY = "YOUR_PINECONE_API_KEY" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="pinecone_agent", instruction="Help users manage and search their Pinecone vector indexes", tools=[ @@ -81,7 +81,7 @@ filtering, and search across multiple indexes with reranking. const PINECONE_API_KEY = "YOUR_PINECONE_API_KEY"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "pinecone_agent", instruction: "Help users manage and search their Pinecone vector indexes", tools: [ diff --git a/docs/integrations/postman.md b/docs/integrations/postman.md index bcd68d27fe..8fae203b63 100644 --- a/docs/integrations/postman.md +++ b/docs/integrations/postman.md @@ -51,7 +51,7 @@ natural language interactions. POSTMAN_API_KEY = "YOUR_POSTMAN_API_KEY" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="postman_agent", instruction="Help users manage their Postman workspaces and collections", tools=[ @@ -87,7 +87,7 @@ natural language interactions. POSTMAN_API_KEY = "YOUR_POSTMAN_API_KEY" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="postman_agent", instruction="Help users manage their Postman workspaces and collections", tools=[ @@ -116,7 +116,7 @@ natural language interactions. const POSTMAN_API_KEY = "YOUR_POSTMAN_API_KEY"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "postman_agent", instruction: "Help users manage their Postman workspaces and collections", tools: [ @@ -150,7 +150,7 @@ natural language interactions. const POSTMAN_API_KEY = "YOUR_POSTMAN_API_KEY"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "postman_agent", instruction: "Help users manage their Postman workspaces and collections", tools: [ diff --git a/docs/integrations/qdrant.md b/docs/integrations/qdrant.md index d1ed2e7bfb..50bf7e9d61 100644 --- a/docs/integrations/qdrant.md +++ b/docs/integrations/qdrant.md @@ -50,7 +50,7 @@ retrieve information using semantic search. # QDRANT_API_KEY = "YOUR_QDRANT_API_KEY" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="qdrant_agent", instruction="Help users store and retrieve information using semantic search", tools=[ @@ -84,7 +84,7 @@ retrieve information using semantic search. // const QDRANT_API_KEY = "YOUR_QDRANT_API_KEY"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "qdrant_agent", instruction: "Help users store and retrieve information using semantic search", tools: [ diff --git a/docs/integrations/stripe.md b/docs/integrations/stripe.md index f3ec7c21d5..fed93bfc21 100644 --- a/docs/integrations/stripe.md +++ b/docs/integrations/stripe.md @@ -49,7 +49,7 @@ operations. STRIPE_SECRET_KEY = "YOUR_STRIPE_SECRET_KEY" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="stripe_agent", instruction="Help users manage their Stripe account", tools=[ @@ -85,7 +85,7 @@ operations. STRIPE_SECRET_KEY = "YOUR_STRIPE_SECRET_KEY" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="stripe_agent", instruction="Help users manage their Stripe account", tools=[ @@ -111,7 +111,7 @@ operations. const STRIPE_SECRET_KEY = "YOUR_STRIPE_SECRET_KEY"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "stripe_agent", instruction: "Help users manage their Stripe account", tools: [ @@ -145,7 +145,7 @@ operations. const STRIPE_SECRET_KEY = "YOUR_STRIPE_SECRET_KEY"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "stripe_agent", instruction: "Help users manage their Stripe account", tools: [ diff --git a/docs/integrations/supermetrics.md b/docs/integrations/supermetrics.md index c3fc9257ce..c82cc4a060 100644 --- a/docs/integrations/supermetrics.md +++ b/docs/integrations/supermetrics.md @@ -55,7 +55,7 @@ accounts using natural language. SUPERMETRICS_API_KEY = "YOUR_SUPERMETRICS_API_KEY" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="supermetrics_agent", instruction="Help users query and analyze their marketing data from Supermetrics", tools=[ @@ -81,7 +81,7 @@ accounts using natural language. const SUPERMETRICS_API_KEY = "YOUR_SUPERMETRICS_API_KEY"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "supermetrics_agent", instruction: "Help users query and analyze their marketing data from Supermetrics", tools: [ diff --git a/docs/integrations/temporal.md b/docs/integrations/temporal.md index 49a6736b14..718a2554c9 100644 --- a/docs/integrations/temporal.md +++ b/docs/integrations/temporal.md @@ -104,7 +104,7 @@ weather_tool = activity_tool( agent = Agent( name="weather_agent", model=TemporalModel( - "gemini-2.5-pro", + "gemini-flash-latest", activity_config=ActivityConfig(summary="Weather Agent")), tools=[weather_tool], ) @@ -230,7 +230,7 @@ client = await Client.connect( # not_in_workflow_toolset lets this agent also run locally with `adk web`. agent = Agent( name="tool_agent", - model=TemporalModel("gemini-2.5-pro"), + model=TemporalModel("gemini-flash-latest"), tools=[TemporalMcpToolSet("my-tools", not_in_workflow_toolset=toolset_factory)], ) ``` diff --git a/docs/integrations/windsor-ai.md b/docs/integrations/windsor-ai.md index e488f764a7..0255892c59 100644 --- a/docs/integrations/windsor-ai.md +++ b/docs/integrations/windsor-ai.md @@ -54,7 +54,7 @@ business data using natural language, without writing SQL or custom scripts. WINDSOR_API_KEY = "YOUR_WINDSOR_API_KEY" root_agent = Agent( - model="gemini-2.5-pro", + model="gemini-flash-latest", name="windsor_agent", instruction="Help users analyze their marketing and business data.", tools=[ @@ -80,7 +80,7 @@ business data using natural language, without writing SQL or custom scripts. const WINDSOR_API_KEY = "YOUR_WINDSOR_API_KEY"; const rootAgent = new LlmAgent({ - model: "gemini-2.5-pro", + model: "gemini-flash-latest", name: "windsor_agent", instruction: "Help users analyze their marketing and business data.", tools: [ diff --git a/docs/sessions/memory.md b/docs/sessions/memory.md index eb6faf33de..81a1599f7d 100644 --- a/docs/sessions/memory.md +++ b/docs/sessions/memory.md @@ -28,12 +28,12 @@ The Python ADK ships three `MemoryService` implementations. Use the table below | **Feature** | **InMemoryMemoryService** | **VertexAiMemoryBankService** | **VertexAiRagMemoryService** | | :--- | :--- | :--- | :--- | -| **Persistence** | None (data is lost on restart) | Yes (Managed by Agent Platform) | Yes (stored in RAG Engine) | +| **Persistence** | None (data is lost on restart) | Yes (Managed by Agent Platform) | Yes (stored in Knowledge Engine) | | **Primary Use Case** | Prototyping, local development, and simple testing. | Building meaningful, evolving memories from user conversations. | Vector-search retrieval over the full conversation corpus, or alongside other RAG-indexed content. | -| **Memory Extraction** | Stores full conversation | Extracts [meaningful information](https://cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/memory-bank/generate-memories) from conversations and consolidates it with existing memories (powered by LLM) | Stores full conversation, indexed by [RAG Engine](https://cloud.google.com/vertex-ai/generative-ai/docs/rag-engine/rag-overview). | -| **Search Capability** | Basic keyword matching. | Advanced semantic search. | Vector similarity search over RAG Engine. | -| **Setup Complexity** | None. It's the default. | Low. Requires an [Agent Runtime](https://cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/memory-bank/overview) instance on Agent Platform. | Medium. Requires [RAG Engine](https://cloud.google.com/vertex-ai/generative-ai/docs/rag-engine/manage-your-rag-corpus). | -| **Dependencies** | None. | Google Cloud Project, Agent Platform API | Google Cloud Project, RAG Engine, the Agent Platform SDK (optional install). | +| **Memory Extraction** | Stores full conversation | Extracts [meaningful information](https://cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/memory-bank/generate-memories) from conversations and consolidates it with existing memories (powered by LLM) | Stores full conversation, indexed by [Knowledge Engine](https://cloud.google.com/vertex-ai/generative-ai/docs/rag-engine/rag-overview). | +| **Search Capability** | Basic keyword matching. | Advanced semantic search. | Vector similarity search over Knowledge Engine. | +| **Setup Complexity** | None. It's the default. | Low. Requires an [Agent Runtime](https://cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/memory-bank/overview) instance on Agent Platform. | Medium. Requires [Knowledge Engine](https://cloud.google.com/vertex-ai/generative-ai/docs/rag-engine/manage-your-rag-corpus). | +| **Dependencies** | None. | Google Cloud Project, Agent Platform API | Google Cloud Project, Knowledge Engine, the Agent Platform SDK (optional install). | | **When to use it** | When you want to search across multiple sessions’ chat histories for prototyping. | When you want your agent to remember and learn from past interactions. | When you already have RAG infrastructure or want to retrieve over raw conversation transcripts. | `VertexAiRagMemoryService` is only exported from `google.adk.memory` when the Agent Platform SDK is installed. Memory Bank and RAG-backed memory are documented in [Memory Bank](#memory-bank) and [RAG Memory](#rag-memory) below. @@ -397,7 +397,7 @@ Or, you can configure your agent to use the Memory Bank by manually instantiatin ## RAG Memory -The `VertexAiRagMemoryService` stores conversations in [RAG Engine](https://cloud.google.com/vertex-ai/generative-ai/docs/rag-engine/rag-overview) and retrieves them by vector similarity. Use it when you already have RAG infrastructure or want raw transcript retrieval rather than the LLM-extracted memories produced by Memory Bank. Requires the Agent Platform SDK. +The `VertexAiRagMemoryService` stores conversations in [Knowledge Engine](https://cloud.google.com/vertex-ai/generative-ai/docs/rag-engine/rag-overview) and retrieves them by vector similarity. Use it when you already have RAG infrastructure or want raw transcript retrieval rather than the LLM-extracted memories produced by Memory Bank. Requires the Agent Platform SDK. === "Python" diff --git a/docs/sessions/session/index.md b/docs/sessions/session/index.md index 1a3921af0a..8e3be1d20f 100644 --- a/docs/sessions/session/index.md +++ b/docs/sessions/session/index.md @@ -210,7 +210,7 @@ the storage backend that best suits your needs: * A Google Cloud project (`pip install vertexai`) * A Google Cloud storage bucket that can be configured by this [step](https://cloud.google.com/vertex-ai/docs/pipelines/configure-project#storage). - * A Reasoning Engine resource name/ID that can setup following this + * An Agent Runtime resource name/ID that can setup following this [tutorial](/deploy/agent-runtime/). * If you do not have a Google Cloud project and you want to try the VertexAiSessionService, see [Agent Platform Express Mode](/integrations/express-mode/). * **Best for:** Scalable production applications deployed on Google Cloud, diff --git a/docs/workflows/dynamic.md b/docs/workflows/dynamic.md index b1417c0509..80f13742c2 100644 --- a/docs/workflows/dynamic.md +++ b/docs/workflows/dynamic.md @@ -310,9 +310,9 @@ async def parallel_supervisor( Dynamic workflows in ADK can also include human input or human in the loop (HITL) steps. You build human input into workflows by yielding a -***RequestInput*** from a `@node` function, which pauses the workflow and waits -for user input. The following code example shows how to build a human input node -and include it in a workflow: +***RequestInput*** from a node, which pauses the workflow and waits for user +input. The following code example shows how to build a human input node and +include it in a workflow: ```python from typing import Any diff --git a/examples/typescript/snippets/skills/get_started.ts b/examples/typescript/snippets/skills/get_started.ts index 6b273bd212..c91f91252a 100644 --- a/examples/typescript/snippets/skills/get_started.ts +++ b/examples/typescript/snippets/skills/get_started.ts @@ -4,7 +4,7 @@ * SPDX-License-Identifier: Apache-2.0 */ -// [START full_example] +// --8<-- [start:full_example] import {Agent, FunctionTool, SkillToolset, loadSkillFromDir} from '@google/adk'; import * as path from 'node:path'; import {z} from 'zod'; @@ -42,4 +42,4 @@ const rootAgent = new Agent({ }); export default rootAgent; -// [END full_example] +// --8<-- [end:full_example] diff --git a/examples/typescript/snippets/skills/inline_skill.ts b/examples/typescript/snippets/skills/inline_skill.ts index 0f64166c50..097bc5d9c4 100644 --- a/examples/typescript/snippets/skills/inline_skill.ts +++ b/examples/typescript/snippets/skills/inline_skill.ts @@ -4,7 +4,7 @@ * SPDX-License-Identifier: Apache-2.0 */ -// [START full_example] +// --8<-- [start:full_example] import {Agent, Skill, SkillToolset} from '@google/adk'; const greetingSkill: Skill = { @@ -33,4 +33,4 @@ const rootAgent = new Agent({ }); export default rootAgent; -// [END full_example] +// --8<-- [end:full_example] diff --git a/mkdocs.yml b/mkdocs.yml index feeadc65ab..97c116dda7 100644 --- a/mkdocs.yml +++ b/mkdocs.yml @@ -199,8 +199,9 @@ plugins: 'tools/third-party/qdrant.md': 'integrations/qdrant.md' 'tools/google-cloud/spanner.md': 'integrations/spanner.md' 'tools/third-party/stripe.md': 'integrations/stripe.md' - 'tools/google-cloud/vertex-ai-rag-engine.md': 'integrations/rag-engine.md' - 'integrations/vertex-ai-rag-engine.md': 'integrations/rag-engine.md' + 'tools/google-cloud/vertex-ai-rag-engine.md': 'integrations/knowledge-engine.md' + 'integrations/vertex-ai-rag-engine.md': 'integrations/knowledge-engine.md' + 'integrations/rag-engine.md': 'integrations/knowledge-engine.md' 'tools/google-cloud/vertex-ai-search.md': 'integrations/agent-search.md' 'integrations/vertex-ai-search.md': 'integrations/agent-search.md' 'grounding/vertex_ai_search_grounding.md': 'grounding/grounding_with_search.md' diff --git a/tools/feature-matrix/start.md b/tools/feature-matrix/start.md index 499e9e2a4c..29e42128c6 100644 --- a/tools/feature-matrix/start.md +++ b/tools/feature-matrix/start.md @@ -22,7 +22,7 @@ The following matrix shows which ADK features are currently supported in each la | | **Built-in: Google Search** | [Supported](https://adk.dev/tools/built-in-tools/#google-search) | [Supported](https://adk.dev/tools/built-in-tools/#google-search) | | | **Built-in: Code Execution** | [Supported](https://adk.dev/tools/built-in-tools/#code-execution) | [Supported](https://adk.dev/tools/built-in-tools/#code-execution) | | | **Built-in: GKE Code Executor** | [Supported](https://adk.dev/tools/built-in-tools/#gke-code-executor) | Not Supported | -| | **Built-in: RAG Engine** | [Supported](https://adk.dev/tools/built-in-tools/#rag-engine) | Not Supported | +| | **Built-in: Knowledge Engine** | [Supported](https://adk.dev/integrations/knowledge-engine/) | Not Supported | | | **Built-in: Agent Search** | [Supported](https://adk.dev/tools/built-in-tools/#agent-search) | Not Supported | | | **Built-in: BigQuery, Spanner, Bigtable** | [Supported](https://adk.dev/tools/built-in-tools/#bigquery) | Not Supported | | | **Google Cloud Tools (API Hub, etc.)** | [Supported](https://adk.dev/tools/google-cloud-tools/#apigee-api-hub-tools) | Not Supported |